Computational Tools for the Analysis, Engineering and Rational Design of Genetic and Metabolic Networks in Single- and Multi-Species Microbial Systems

Metabolism is defined as the full complement of biochemical transformations in living systems required to maintain life. Reactions in a metabolic network are catalyzed and regulated by a complex web of genetic and enzymatic interactions. Efficient engineering and design of natural and synthetic microbial systems for biomedical and biotechnological applications require a system-level understanding of interactions in the gene and metabolic levels and combinations thereof. It is thus timely to develop systems-based analysis tools and modeling techniques capable of providing a quantitative understanding of genetic and metabolic interactions.

Here, I review my recent work on the development of efficient computational and modeling tools to shed light onto the less-understood aspects of metabolism in single- and multi-species microbial systems, and to identify promising network rewiring and/or synthetic genetic systems design strategies for biotechnology applications. I first introduce a workflow that combines the ensemble modeling of metabolic networks with customized optimization-based algorithms for creating large-scale kinetic models of metabolism. This is exemplified by creating a large-scale kinetic model of E. coli using multiple fluxomic and concentration data. I next introduce Kinetic OptForce, a computational strain design protocol integrating the available kinetic information with the genome-scale stoichiometric-based models of metabolism for the system-level identification of the metabolic interventions leading to the targeted overproduction of a biochemical of interest. Results and lessons learned by application of Kinetic OptForce for the overproduction of a number of value-added chemicals and biofuels (e.g., fatty acids, Neurosporene , L-serine etc.) in E. coli and S. cerevisiae will be shared.

Metabolic interventions predicted by OptForce can be experimentally implemented in a variety of ways by using different operon designs, however, a system-level procedure for designing synthetic operons is still lacking. To bridge this gap, I will describe a novel  computational framework, called Operon Calculator,  for rationally designing a library of synthetic operons, which enables pinpointing optimal enzyme expression levels consistent with the predicted metabolic interventions by Kinetic OptForce.  Operon Calculator is an optimization procedure based on a multi-level genetic algorithm that accounts for multiple competing objective functions and uses biophysical models of translation initiation and RNA folding to efficiently search the high dimensional space of genetic parts sequences.

The focus is then shifted from single-species to multi-species microbial systems, where multiple microorganisms communicate through unidirectional or bidirectional exchange of biochemical cues. Here, I introduce a multi-level and multi-objective optimization framework, called OptCom, for the metabolic modeling and analysis of microbial communities using genome-scale metabolic models. Finally, given that microbial communities change with time (e.g., day/night cycle) and location (e.g., nutrient gradients), I will discuss the ongoing efforts to capture the spatio-temporal changes in microbial communities by extension of the OptCom procedure. Overall, the wide array of computational and modeling tools introduced in this poster highlight their utility for the model-driven analysis and redesign of biological networks to aid metabolic engineering efforts and complement experimental studies.